[PATCH v2 0/4] mm: move vma_(kernel|mmu)_pagesize() out of hugetlb.c

[PATCH v2 0/4] mm: move vma_(kernel|mmu)_pagesize() out of hugetlb.c

[David Hildenbrand (Arm)]

Looking into vma_(kernel|mmu)_pagesize(), I realized that there is one
scenario where DAX would not do the right thing when the kernel is
not compiled with hugetlb support.

Without hugetlb support, vma_(kernel|mmu)_pagesize() will always return
PAGE_SIZE instead of using the ->pagesize() result provided by dax-device
code.

Fix that by moving vma_kernel_pagesize() to core MM code, where it belongs.
I don't think this is stable material, but am not 100% sure.

Also, move vma_mmu_pagesize() while at it. Remove the unnecessary hugetlb.h
inclusion from KVM code.

Cross-compiled heavily.

v1 -> v2:
* Fixup VMA test in patch #2
* Actually CC all people listed below on all patches.

Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: "Christophe Leroy (CS GROUP)" <chleroy@kernel.org>
Cc: Muchun Song <muchun.song@linux.dev>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Lorenzo Stoakes <ljs@kernel.org>
Cc: "Liam R. Howlett" <Liam.Howlett@oracle.com>
Cc: Vlastimil Babka <vbabka@kernel.org>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Jann Horn <jannh@google.com>
Cc: Pedro Falcato <pfalcato@suse.de>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Dan Williams <dan.j.williams@intel.com>

David Hildenbrand (Arm) (4):
  mm: move vma_kernel_pagesize() from hugetlb to mm.h
  mm: move vma_mmu_pagesize() from hugetlb to vma.c
  KVM: remove hugetlb.h inclusion
  KVM: PPC: remove hugetlb.h inclusion

 arch/powerpc/kvm/book3s_hv.c       |  1 -
 include/linux/hugetlb.h            | 14 --------------
 include/linux/mm.h                 | 22 ++++++++++++++++++++++
 mm/hugetlb.c                       | 28 ----------------------------
 mm/vma.c                           | 21 +++++++++++++++++++++
 tools/testing/vma/include/custom.h |  5 +++++
 virt/kvm/kvm_main.c                |  1 -
 7 files changed, 48 insertions(+), 44 deletions(-)


base-commit: f75825cdfc4c5477cffcfd2cafa4e5ce5aa67f13
-- 
2.43.0

[PATCH v2 1/4] mm: move vma_kernel_pagesize() from hugetlb to mm.h

[David Hildenbrand (Arm)]

In the past, only hugetlb had special "vma_kernel_pagesize()"
requirements, so it provided its own implementation.

In commit 05ea88608d4e ("mm, hugetlbfs: introduce ->pagesize() to
vm_operations_struct") we generalized that approach by providing a
vm_ops->pagesize() callback to be used by device-dax.

Once device-dax started using that callback in commit c1d53b92b95c
("device-dax: implement ->pagesize() for smaps to report MMUPageSize")
it was missed that CONFIG_DEV_DAX does not depend on hugetlb support.

So building a kernel with CONFIG_DEV_DAX but without CONFIG_HUGETLBFS
would not pick up that value.

Fix it by moving vma_kernel_pagesize() to mm.h, providing only a single
implementation. While at it, improve the kerneldoc a bit.

Ideally, we'd move vma_mmu_pagesize() as well to the header. However,
its __weak symbol might be overwritten by a PPC variant in hugetlb code.
So let's leave it in there for now, as it really only matters for some
hugetlb oddities.

This was found by code inspection.

Fixes: c1d53b92b95c ("device-dax: implement ->pagesize() for smaps to report MMUPageSize")
Reviewed-by: Lorenzo Stoakes (Oracle) <ljs@kernel.org>
Acked-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: David Hildenbrand (Arm) <david@kernel.org>
---
 include/linux/hugetlb.h |  7 -------
 include/linux/mm.h      | 20 ++++++++++++++++++++
 mm/hugetlb.c            | 17 -----------------
 3 files changed, 20 insertions(+), 24 deletions(-)

diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
index 65910437be1c..44c1848a2c21 100644
--- a/include/linux/hugetlb.h
+++ b/include/linux/hugetlb.h
@@ -777,8 +777,6 @@ static inline unsigned long huge_page_size(const struct hstate *h)
 	return (unsigned long)PAGE_SIZE << h->order;
 }
 
-extern unsigned long vma_kernel_pagesize(struct vm_area_struct *vma);
-
 extern unsigned long vma_mmu_pagesize(struct vm_area_struct *vma);
 
 static inline unsigned long huge_page_mask(struct hstate *h)
@@ -1177,11 +1175,6 @@ static inline unsigned long huge_page_mask(struct hstate *h)
 	return PAGE_MASK;
 }
 
-static inline unsigned long vma_kernel_pagesize(struct vm_area_struct *vma)
-{
-	return PAGE_SIZE;
-}
-
 static inline unsigned long vma_mmu_pagesize(struct vm_area_struct *vma)
 {
 	return PAGE_SIZE;
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 44e04a42fe77..227809790f1a 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -1307,6 +1307,26 @@ static inline bool vma_is_shared_maywrite(const struct vm_area_struct *vma)
 	return is_shared_maywrite(&vma->flags);
 }
 
+/**
+ * vma_kernel_pagesize - Default page size granularity for this VMA.
+ * @vma: The user mapping.
+ *
+ * The kernel page size specifies in which granularity VMA modifications
+ * can be performed. Folios in this VMA will be aligned to, and at least
+ * the size of the number of bytes returned by this function.
+ *
+ * The default kernel page size is not affected by Transparent Huge Pages
+ * being in effect.
+ *
+ * Return: The default page size granularity for this VMA.
+ */
+static inline unsigned long vma_kernel_pagesize(struct vm_area_struct *vma)
+{
+	if (unlikely(vma->vm_ops && vma->vm_ops->pagesize))
+		return vma->vm_ops->pagesize(vma);
+	return PAGE_SIZE;
+}
+
 static inline
 struct vm_area_struct *vma_find(struct vma_iterator *vmi, unsigned long max)
 {
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 1d41fa3dd43e..66eadfa9e958 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -1017,23 +1017,6 @@ static pgoff_t vma_hugecache_offset(struct hstate *h,
 			(vma->vm_pgoff >> huge_page_order(h));
 }
 
-/**
- * vma_kernel_pagesize - Page size granularity for this VMA.
- * @vma: The user mapping.
- *
- * Folios in this VMA will be aligned to, and at least the size of the
- * number of bytes returned by this function.
- *
- * Return: The default size of the folios allocated when backing a VMA.
- */
-unsigned long vma_kernel_pagesize(struct vm_area_struct *vma)
-{
-	if (vma->vm_ops && vma->vm_ops->pagesize)
-		return vma->vm_ops->pagesize(vma);
-	return PAGE_SIZE;
-}
-EXPORT_SYMBOL_GPL(vma_kernel_pagesize);
-
 /*
  * Return the page size being used by the MMU to back a VMA. In the majority
  * of cases, the page size used by the kernel matches the MMU size. On
-- 
2.43.0

[PATCH v2 2/4] mm: move vma_mmu_pagesize() from hugetlb to vma.c

[David Hildenbrand (Arm)]

vma_mmu_pagesize() is also queried on non-hugetlb VMAs and does not
really belong into hugetlb.c.

PPC64 provides a custom overwrite with CONFIG_HUGETLB_PAGE, see
arch/powerpc/mm/book3s64/slice.c, so we cannot easily make this a
static inline function.

So let's move it to vma.c and add some proper kerneldoc.

To make vma tests happy, add a simple vma_kernel_pagesize() stub in
tools/testing/vma/include/custom.h.

Reviewed-by: Lorenzo Stoakes (Oracle) <ljs@kernel.org>
Acked-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Signed-off-by: David Hildenbrand (Arm) <david@kernel.org>
---
 include/linux/hugetlb.h            |  7 -------
 include/linux/mm.h                 |  2 ++
 mm/hugetlb.c                       | 11 -----------
 mm/vma.c                           | 21 +++++++++++++++++++++
 tools/testing/vma/include/custom.h |  5 +++++
 5 files changed, 28 insertions(+), 18 deletions(-)

diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
index 44c1848a2c21..aaf3d472e6b5 100644
--- a/include/linux/hugetlb.h
+++ b/include/linux/hugetlb.h
@@ -777,8 +777,6 @@ static inline unsigned long huge_page_size(const struct hstate *h)
 	return (unsigned long)PAGE_SIZE << h->order;
 }
 
-extern unsigned long vma_mmu_pagesize(struct vm_area_struct *vma);
-
 static inline unsigned long huge_page_mask(struct hstate *h)
 {
 	return h->mask;
@@ -1175,11 +1173,6 @@ static inline unsigned long huge_page_mask(struct hstate *h)
 	return PAGE_MASK;
 }
 
-static inline unsigned long vma_mmu_pagesize(struct vm_area_struct *vma)
-{
-	return PAGE_SIZE;
-}
-
 static inline unsigned int huge_page_order(struct hstate *h)
 {
 	return 0;
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 227809790f1a..22d338933c84 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -1327,6 +1327,8 @@ static inline unsigned long vma_kernel_pagesize(struct vm_area_struct *vma)
 	return PAGE_SIZE;
 }
 
+unsigned long vma_mmu_pagesize(struct vm_area_struct *vma);
+
 static inline
 struct vm_area_struct *vma_find(struct vma_iterator *vmi, unsigned long max)
 {
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 66eadfa9e958..f6ecca9aae01 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -1017,17 +1017,6 @@ static pgoff_t vma_hugecache_offset(struct hstate *h,
 			(vma->vm_pgoff >> huge_page_order(h));
 }
 
-/*
- * Return the page size being used by the MMU to back a VMA. In the majority
- * of cases, the page size used by the kernel matches the MMU size. On
- * architectures where it differs, an architecture-specific 'strong'
- * version of this symbol is required.
- */
-__weak unsigned long vma_mmu_pagesize(struct vm_area_struct *vma)
-{
-	return vma_kernel_pagesize(vma);
-}
-
 /*
  * Flags for MAP_PRIVATE reservations.  These are stored in the bottom
  * bits of the reservation map pointer, which are always clear due to
diff --git a/mm/vma.c b/mm/vma.c
index be64f781a3aa..e95fd5a5fe5c 100644
--- a/mm/vma.c
+++ b/mm/vma.c
@@ -3300,3 +3300,24 @@ int insert_vm_struct(struct mm_struct *mm, struct vm_area_struct *vma)
 
 	return 0;
 }
+
+/**
+ * vma_mmu_pagesize - Default MMU page size granularity for this VMA.
+ * @vma: The user mapping.
+ *
+ * In the common case, the default page size used by the MMU matches the
+ * default page size used by the kernel (see vma_kernel_pagesize()). On
+ * architectures where it differs, an architecture-specific 'strong' version
+ * of this symbol is required.
+ *
+ * The default MMU page size is not affected by Transparent Huge Pages
+ * being in effect, or any usage of larger MMU page sizes (either through
+ * architectural huge-page mappings or other explicit/implicit coalescing of
+ * virtual ranges performed by the MMU).
+ *
+ * Return: The default MMU page size granularity for this VMA.
+ */
+__weak unsigned long vma_mmu_pagesize(struct vm_area_struct *vma)
+{
+	return vma_kernel_pagesize(vma);
+}
diff --git a/tools/testing/vma/include/custom.h b/tools/testing/vma/include/custom.h
index 802a76317245..4305a5b6e433 100644
--- a/tools/testing/vma/include/custom.h
+++ b/tools/testing/vma/include/custom.h
@@ -117,3 +117,8 @@ static inline vma_flags_t __mk_vma_flags(size_t count, const vma_flag_t *bits)
 			vma_flag_set(&flags, bits[i]);
 	return flags;
 }
+
+static inline unsigned long vma_kernel_pagesize(struct vm_area_struct *vma)
+{
+	return PAGE_SIZE;
+}
-- 
2.43.0

[PATCH v2 3/4] KVM: remove hugetlb.h inclusion

[David Hildenbrand (Arm)]

hugetlb.h is no longer required now that we moved vma_kernel_pagesize()
to mm.h.

Reviewed-by: Lorenzo Stoakes (Oracle) <ljs@kernel.org>
Acked-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Signed-off-by: David Hildenbrand (Arm) <david@kernel.org>
---
 virt/kvm/kvm_main.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 22f8a672e1fd..58059648b881 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -41,7 +41,6 @@
 #include <linux/spinlock.h>
 #include <linux/compat.h>
 #include <linux/srcu.h>
-#include <linux/hugetlb.h>
 #include <linux/slab.h>
 #include <linux/sort.h>
 #include <linux/bsearch.h>
-- 
2.43.0

[PATCH v2 4/4] KVM: PPC: remove hugetlb.h inclusion

[David Hildenbrand (Arm)]

hugetlb.h is no longer required now that we moved vma_kernel_pagesize()
to mm.h.

Reviewed-by: Lorenzo Stoakes (Oracle) <ljs@kernel.org>
Acked-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Signed-off-by: David Hildenbrand (Arm) <david@kernel.org>
---
 arch/powerpc/kvm/book3s_hv.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/arch/powerpc/kvm/book3s_hv.c b/arch/powerpc/kvm/book3s_hv.c
index 08e5816fdd61..61dbeea317f3 100644
--- a/arch/powerpc/kvm/book3s_hv.c
+++ b/arch/powerpc/kvm/book3s_hv.c
@@ -36,7 +36,6 @@
 #include <linux/gfp.h>
 #include <linux/vmalloc.h>
 #include <linux/highmem.h>
-#include <linux/hugetlb.h>
 #include <linux/kvm_irqfd.h>
 #include <linux/irqbypass.h>
 #include <linux/module.h>
-- 
2.43.0

Re: [PATCH v2 4/4] KVM: PPC: remove hugetlb.h inclusion

[Ritesh Harjani]

"David Hildenbrand (Arm)" <david@kernel.org> writes:

> hugetlb.h is no longer required now that we moved vma_kernel_pagesize()
> to mm.h.
>
> Reviewed-by: Lorenzo Stoakes (Oracle) <ljs@kernel.org>
> Acked-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
> Signed-off-by: David Hildenbrand (Arm) <david@kernel.org>
> ---
>  arch/powerpc/kvm/book3s_hv.c | 1 -

Make sense to me. Please feel free to add:

Reviewed-by: Ritesh Harjani <ritesh.list@gmail.com>


>  1 file changed, 1 deletion(-)
>
> diff --git a/arch/powerpc/kvm/book3s_hv.c b/arch/powerpc/kvm/book3s_hv.c
> index 08e5816fdd61..61dbeea317f3 100644
> --- a/arch/powerpc/kvm/book3s_hv.c
> +++ b/arch/powerpc/kvm/book3s_hv.c
> @@ -36,7 +36,6 @@
>  #include <linux/gfp.h>
>  #include <linux/vmalloc.h>
>  #include <linux/highmem.h>
> -#include <linux/hugetlb.h>
>  #include <linux/kvm_irqfd.h>
>  #include <linux/irqbypass.h>
>  #include <linux/module.h>
> -- 
> 2.43.0